Validation is obtained or achieved after passing a certain test.
Validation is the action and effect of validating ( making something valid , giving it strength or firmness ). The adjective valid, on the other hand, refers to that which has legal weight or that is rigid and subsistent.
For example: "We have tried to verify the authenticity of the product , but the truth is that it did not pass the validation process" , "The owner has already validated the project, which will be developed in the coming months" , "The program did not pass the validation process and, therefore, stopped working .
Software validation
In the field of software creation, validation testing is known as the review process to which a computer program is subjected to verify that it meets its specifications. This, which usually takes place at the end of the development stage, is carried out mainly with the intention of confirming that the application can carry out the tasks that its potential users expect of it.
Validation tests are also carried out to determine if a software license is legal or if it is a counterfeit (a pirated copy). Some versions of the Windows operating system perform these validation tests automatically (without the user requiring it). When it happens that the process is not passed, the system itself warns the user that they could be a victim of counterfeiting.
The approval method depends on the discipline and context.
Methods used in chemistry
Taking as an example the field of analytical chemistry , which studies the composition of materials through laboratory tests, it is known that it is possible to use a validation method for a given analyte (the element of interest in a sample), using a certain instrumentation, of the sample in question and carrying out a specific treatment of the data, and that said method can be applied in various laboratories with equivalent results, as long as they meet the same equipment and personnel requirements.
There are different validation methods , which must be used after the development and optimization stage:
* Blind method : Using samples of known concentration of a particular compound, analysts can determine whether they meet a specific set of requirements. Although this method depends largely on those who carry it out, it is a low-complexity practice that requires little time and guarantees the impartiality of its participants. Three modalities can be distinguished:
+ zero blind : only one person intervenes;
+ single blind : carried out by two analysts;
+ double blind : three professionals participate, dividing the work in a very specific way. The first analyst is in charge of preparing the samples and carries out, together with the second, the relevant analyses . The last one has the task of comparing the results without knowing who owns each one.
* Validation with reference materials : it is based on a material standard or a sample that has been authenticated and the results obtained with them; The condition is that there is an absolute match after validation. It is worth mentioning that these materials are distributed by various laboratories.
* Inter-laboratory comparison : this is the most used method, both for the validation of a method and for the preparation of reference materials. It entails, however, a considerable economic and time investment.
* Comparison with an accepted method : similar to the last two, it consists of contrasting the results obtained by two particular validations, using any of the three modalities of the blind method.
Validation in statistics
Cross validation , finally, is a statistical practice.
This procedure consists of fragmenting a data sample into subsets to analyze one of them and then validating said analysis with the rest of the subsets.